Cytomegalovirus (CMV) is a ubiquitous herpesvirus that establishes a systemic, persistent infection. CMV has evolved to rarely cause serious disease in humans because systemic, life-long immune surveillance keeps the virus in check. In fact, CMV stimulates the largest known T cell populations in the circulation of humans. These T cells accumulate over time in a process called memory inflation and migrate systemically to control CMV by shutting down viral reactivation from latency. For these reasons, CMV may serve as a tool for new vaccines against diseases such as cancer and HIV. However, much remains unknown about how CMV manages to successfully reactivate to be periodically shed from a host, in the midst of immune surveillance. CMV can cause devastating disease in a developing fetus when the virus is transmitted to a pregnant woman. Thus, a vaccine to prevent CMV transmission is rated as a highest priority by the Institute of Medicine. Understanding immune surveillance at sites of viral shedding will be key to preventing transmission and CMV disease. Recent work has shown that a T cell population called resident memory T cells (TRM) are established at sites in the body that may face viral reactivation. Indeed, TRM cells may help control herpesvirus reactivation. However, there have been no studies of CMV-specific TRM cells. Using the natural mouse herpesvirus, murine (M)CMV, our data show that many MCMV-specifics TRM cells developed in the salivary and mammary glands - two sites from which MCMV is known to be shed. The ontogeny of TRM cells is poorly defined, and this gap is critical because these T cells are best positioned and possibly critical for controlling herpesvirus reactivation. Moreover, the promotion of such first responders - cells positioned at the site of pathogen invasion - is the major advantage of CMV-vectored vaccines. Aim 1 of the current proposal will characterize MCMV-specific TRM cells, determine whether they localize to infected epithelium, and whether their maintenance depends on known players - namely TGF-, CD103 (E7 integrin) and CD69. Aim 2 will determine whether the development of MCMV-specific TRM cells depends on local viral replication or the efficacy of MCMV immune evasion. MCMV undergoes prolonged replication in the salivary gland and is readily reactivated there, which is typical of herpesviruses at sites of shedding. However, MCMV immune evasion is thought to be highly effective in the salivary gland. Thus, we will determine whether MCMV-specific TRM cells develop because of the local MCMV infection or in spite of it. Finally, our preliminary data show that memory inflation in circulation is driven by a competition for viral antigen. T cells that successfully compete, inflate; those that fail to compete do not. Remarkably, our data suggest that MCMV-specific TRM cells were enriched for T cells that do not undergo memory inflation. Thus, Aim 3, will determine whether T cells that fail to compete for MCMV antigen are preferentially enriched in the TRM pool and whether the affinity of the T cell receptor for viral peptide dictates the decision to join the inflationary or resident T cell pools.